def append_dsl(self, search_results_object, permitted_nodegroups, include_provisional): nested_agg = NestedAgg(path='points', name='geo_aggs') nested_agg_filter = FiltersAgg(name='inner') geo_agg_filter = Bool() if include_provisional is True: geo_agg_filter.filter( Terms(field='points.provisional', terms=['false', 'true'])) else: if include_provisional is False: geo_agg_filter.filter( Terms(field='points.provisional', terms=['false'])) elif include_provisional is 'only provisional': geo_agg_filter.filter( Terms(field='points.provisional', terms=['true'])) geo_agg_filter.filter( Terms(field='points.nodegroup_id', terms=permitted_nodegroups)) nested_agg_filter.add_filter(geo_agg_filter) nested_agg_filter.add_aggregation( GeoHashGridAgg(field='points.point', name='grid', precision=settings.HEX_BIN_PRECISION)) nested_agg_filter.add_aggregation( GeoBoundsAgg(field='points.point', name='bounds')) nested_agg.add_aggregation(nested_agg_filter) search_results_object['query'].add_aggregation(nested_agg)
def get_relations(resourceinstanceid, start, limit, resourceinstance_graphid=None): query = Query(se, start=start, limit=limit) bool_filter = Bool() bool_filter.should( Terms(field="resourceinstanceidfrom", terms=resourceinstanceid)) bool_filter.should( Terms(field="resourceinstanceidto", terms=resourceinstanceid)) if resourceinstance_graphid: graph_id_filter = Bool() graph_id_filter.should( Terms(field="resourceinstancefrom_graphid", terms=resourceinstance_graphid)) graph_id_filter.should( Terms(field="resourceinstanceto_graphid", terms=resourceinstance_graphid)) bool_filter.must(graph_id_filter) query.add_query(bool_filter) return query.search(index=RESOURCE_RELATIONS_INDEX)
def gen_range_agg(gte=None, lte=None, permitted_nodegroups=None): date_query = Bool() date_query.filter( Range(field='dates.date', gte=gte, lte=lte, relation='intersects')) if permitted_nodegroups: date_query.filter( Terms(field='dates.nodegroup_id', terms=permitted_nodegroups)) date_ranges_query = Bool() date_ranges_query.filter( Range(field='date_ranges.date_range', gte=gte, lte=lte, relation='intersects')) if permitted_nodegroups: date_ranges_query.filter( Terms(field='date_ranges.nodegroup_id', terms=permitted_nodegroups)) wrapper_query = Bool() wrapper_query.should( Nested(path='date_ranges', query=date_ranges_query)) wrapper_query.should(Nested(path='dates', query=date_query)) return wrapper_query
def get_relations(resourceinstanceid, start, limit): query = Query(se, start=start, limit=limit) bool_filter = Bool() bool_filter.should(Terms(field="resourceinstanceidfrom", terms=resourceinstanceid)) bool_filter.should(Terms(field="resourceinstanceidto", terms=resourceinstanceid)) query.add_query(bool_filter) return query.search(index="resource_relations")
def gen_range_agg(gte=None, lte=None, permitted_nodegroups=None): date_query = Bool() date_query.filter( Range(field="dates.date", gte=gte, lte=lte, relation="intersects")) if permitted_nodegroups is not None: date_query.filter( Terms(field="dates.nodegroup_id", terms=permitted_nodegroups)) date_ranges_query = Bool() date_ranges_query.filter( Range(field="date_ranges.date_range", gte=gte, lte=lte, relation="intersects")) if permitted_nodegroups is not None: date_ranges_query.filter( Terms(field="date_ranges.nodegroup_id", terms=permitted_nodegroups)) wrapper_query = Bool() wrapper_query.should( Nested(path="date_ranges", query=date_ranges_query)) wrapper_query.should(Nested(path="dates", query=date_query)) return wrapper_query
def append_dsl(self, search_results_object, permitted_nodegroups, include_provisional): nested_agg = NestedAgg(path="points", name="geo_aggs") nested_agg_filter = FiltersAgg(name="inner") geo_agg_filter = Bool() if include_provisional is True: geo_agg_filter.filter( Terms(field="points.provisional", terms=["false", "true"])) else: if include_provisional is False: geo_agg_filter.filter( Terms(field="points.provisional", terms=["false"])) elif include_provisional is "only provisional": geo_agg_filter.filter( Terms(field="points.provisional", terms=["true"])) geo_agg_filter.filter( Terms(field="points.nodegroup_id", terms=permitted_nodegroups)) nested_agg_filter.add_filter(geo_agg_filter) nested_agg_filter.add_aggregation( GeoHashGridAgg(field="points.point", name="grid", precision=settings.HEX_BIN_PRECISION)) nested_agg_filter.add_aggregation( GeoBoundsAgg(field="points.point", name="bounds")) nested_agg.add_aggregation(nested_agg_filter) search_results_object["query"].add_aggregation(nested_agg)
def get_related_resources(resourceid, lang, limit=1000, start=0): ret = {'resource_relationships': [], 'related_resources': []} se = SearchEngineFactory().create() query = Query(se, limit=limit, start=start) query.add_filter(Terms(field='entityid1', terms=resourceid).dsl, operator='or') query.add_filter(Terms(field='entityid2', terms=resourceid).dsl, operator='or') resource_relations = query.search(index='resource_relations', doc_type='all') ret['total'] = resource_relations['hits']['total'] entityids = set() for relation in resource_relations['hits']['hits']: relation['_source']['preflabel'] = get_preflabel_from_valueid( relation['_source']['relationshiptype'], lang) ret['resource_relationships'].append(relation['_source']) entityids.add(relation['_source']['entityid1']) entityids.add(relation['_source']['entityid2']) if len(entityids) > 0: entityids.remove(resourceid) related_resources = se.search(index='entity', doc_type='_all', id=list(entityids)) if related_resources: for resource in related_resources['docs']: ret['related_resources'].append(resource['_source']) return ret
def append_dsl(self, search_results_object, permitted_nodegroups, include_provisional): search_query = Bool() querysting_params = self.request.GET.get(details["componentname"], "") spatial_filter = JSONDeserializer().deserialize(querysting_params) if "features" in spatial_filter: if len(spatial_filter["features"]) > 0: feature_geom = spatial_filter["features"][0]["geometry"] feature_properties = {} if "properties" in spatial_filter["features"][0]: feature_properties = spatial_filter["features"][0][ "properties"] buffer = {"width": 0, "unit": "ft"} if "buffer" in feature_properties: buffer = feature_properties["buffer"] search_buffer = _buffer(feature_geom, buffer["width"], buffer["unit"]) feature_geom = JSONDeserializer().deserialize( search_buffer.geojson) geoshape = GeoShape(field="geometries.geom.features.geometry", type=feature_geom["type"], coordinates=feature_geom["coordinates"]) invert_spatial_search = False if "inverted" in feature_properties: invert_spatial_search = feature_properties["inverted"] spatial_query = Bool() if invert_spatial_search is True: spatial_query.must_not(geoshape) else: spatial_query.filter(geoshape) # get the nodegroup_ids that the user has permission to search spatial_query.filter( Terms(field="geometries.nodegroup_id", terms=permitted_nodegroups)) if include_provisional is False: spatial_query.filter( Terms(field="geometries.provisional", terms=["false"])) elif include_provisional == "only provisional": spatial_query.filter( Terms(field="geometries.provisional", terms=["true"])) search_query.filter( Nested(path="geometries", query=spatial_query)) search_results_object["query"].add_query(search_query) if details["componentname"] not in search_results_object: search_results_object[details["componentname"]] = {} try: search_results_object[ details["componentname"]]["search_buffer"] = feature_geom except NameError: logger.info(_("Feature geometry is not defined"))
def get_relations(resourceinstanceid, start, limit): query = Query(se, limit=limit, start=start) bool_filter = Bool() bool_filter.should( Terms(field='resourceinstanceidfrom', terms=resourceinstanceid)) bool_filter.should( Terms(field='resourceinstanceidto', terms=resourceinstanceid)) query.add_query(bool_filter) return query.search(index='resource_relations', doc_type='all')
def append_dsl(self, search_results_object, permitted_nodegroups, include_provisional): search_query = Bool() querysting_params = self.request.GET.get(details['componentname'], '') spatial_filter = JSONDeserializer().deserialize(querysting_params) if 'features' in spatial_filter: if len(spatial_filter['features']) > 0: feature_geom = spatial_filter['features'][0]['geometry'] feature_properties = {} if 'properties' in spatial_filter['features'][0]: feature_properties = spatial_filter['features'][0][ 'properties'] buffer = {'width': 0, 'unit': 'ft'} if 'buffer' in feature_properties: buffer = feature_properties['buffer'] search_buffer = _buffer(feature_geom, buffer['width'], buffer['unit']) feature_geom = JSONDeserializer().deserialize( search_buffer.json) geoshape = GeoShape(field='geometries.geom.features.geometry', type=feature_geom['type'], coordinates=feature_geom['coordinates']) invert_spatial_search = False if 'inverted' in feature_properties: invert_spatial_search = feature_properties['inverted'] spatial_query = Bool() if invert_spatial_search is True: spatial_query.must_not(geoshape) else: spatial_query.filter(geoshape) # get the nodegroup_ids that the user has permission to search spatial_query.filter( Terms(field='geometries.nodegroup_id', terms=permitted_nodegroups)) if include_provisional is False: spatial_query.filter( Terms(field='geometries.provisional', terms=['false'])) elif include_provisional == 'only provisional': spatial_query.filter( Terms(field='geometries.provisional', terms=['true'])) search_query.filter( Nested(path='geometries', query=spatial_query)) search_results_object['query'].add_query(search_query) if details['componentname'] not in search_results_object: search_results_object[details['componentname']] = {} search_results_object[ details['componentname']]['search_buffer'] = search_buffer.geojson
def delete_index(self, resourceinstanceid=None): """ Deletes all references to a resource from all indexes Keyword Arguments: resourceinstanceid -- the resource instance id to delete from related indexes, if supplied will use this over self.resourceinstanceid """ if resourceinstanceid is None: resourceinstanceid = self.resourceinstanceid resourceinstanceid = str(resourceinstanceid) # delete any related terms query = Query(se) bool_query = Bool() bool_query.filter( Terms(field="resourceinstanceid", terms=[resourceinstanceid])) query.add_query(bool_query) query.delete(index=TERMS_INDEX) # delete any related resource index entries query = Query(se) bool_query = Bool() bool_query.should( Terms(field="resourceinstanceidto", terms=[resourceinstanceid])) bool_query.should( Terms(field="resourceinstanceidfrom", terms=[resourceinstanceid])) query.add_query(bool_query) query.delete(index=RESOURCE_RELATIONS_INDEX) # reindex any related resources query = Query(se) bool_query = Bool() bool_query.filter( Nested(path="ids", query=Terms(field="ids.id", terms=[resourceinstanceid]))) query.add_query(bool_query) results = query.search(index=RESOURCES_INDEX)["hits"]["hits"] for result in results: try: res = Resource.objects.get(pk=result["_id"]) res.load_tiles() res.index() except ObjectDoesNotExist: pass # delete resource index se.delete(index=RESOURCES_INDEX, id=resourceinstanceid) # delete resources from custom indexes for index in settings.ELASTICSEARCH_CUSTOM_INDEXES: es_index = import_class_from_string(index["module"])(index["name"]) es_index.delete_resources(resources=self)
def get_preflabel_from_conceptid(conceptid, lang): ret = None default = { "category": "", "conceptid": "", "language": "", "value": "", "type": "", "id": "" } se = SearchEngineFactory().create() query = Query(se) terms = Terms(field='conceptid', terms=[conceptid]) # Uncomment the following line only after having reindexed ElasticSearch cause currently the Arabic labels are indexed as altLabels # match = Match(field='type', query='prefLabel', type='phrase') query.add_filter(terms) # Uncomment the following line only after having reindexed ElasticSearch cause currently the Arabic labels are indexed as altLabels # query.add_query(match) preflabels = query.search(index='concept_labels')['hits']['hits'] for preflabel in preflabels: # print 'Language at this point %s and label language %s and ret is %s' % (lang, preflabel['_source']['language'], ret) default = preflabel['_source'] # get the label in the preferred language, otherwise get the label in the default language if preflabel['_source']['language'] == lang: # print 'prefLabel from Conceptid: %s' % preflabel['_source'] return preflabel['_source'] if preflabel['_source']['language'].split('-')[0] == lang.split( '-')[0]: ret = preflabel['_source'] if preflabel['_source']['language'] == lang and ret == None: ret = preflabel['_source'] return default if ret == None else ret
def get_preflabel_from_conceptid(conceptid, lang): ret = None default = { "category": "", "conceptid": "", "language": "", "value": "", "type": "", "id": "" } query = Query(se) bool_query = Bool() bool_query.must(Match(field="type", query="prefLabel", type="phrase")) bool_query.filter(Terms(field="conceptid", terms=[conceptid])) query.add_query(bool_query) preflabels = query.search(index=CONCEPTS_INDEX)["hits"]["hits"] for preflabel in preflabels: default = preflabel["_source"] # get the label in the preferred language, otherwise get the label in the default language if preflabel["_source"]["language"] == lang: return preflabel["_source"] if preflabel["_source"]["language"].split("-")[0] == lang.split( "-")[0]: ret = preflabel["_source"] if preflabel["_source"][ "language"] == settings.LANGUAGE_CODE and ret is None: ret = preflabel["_source"] return default if ret is None else ret
def search_terms(request): lang = request.GET.get('lang', settings.LANGUAGE_CODE) se = SearchEngineFactory().create() searchString = request.GET.get('q', '') user_is_reviewer = request.user.groups.filter(name='Resource Reviewer').exists() i = 0 ret = {} for index in ['terms', 'concepts']: query = Query(se, start=0, limit=0) boolquery = Bool() boolquery.should(Match(field='value', query=searchString.lower(), type='phrase_prefix')) boolquery.should(Match(field='value.folded', query=searchString.lower(), type='phrase_prefix')) boolquery.should(Match(field='value.folded', query=searchString.lower(), fuzziness='AUTO', prefix_length=settings.SEARCH_TERM_SENSITIVITY)) if user_is_reviewer is False and index == 'terms': boolquery.filter(Terms(field='provisional', terms=['false'])) query.add_query(boolquery) base_agg = Aggregation(name='value_agg', type='terms', field='value.raw', size=settings.SEARCH_DROPDOWN_LENGTH, order={"max_score": "desc"}) nodegroupid_agg = Aggregation(name='nodegroupid', type='terms', field='nodegroupid') top_concept_agg = Aggregation(name='top_concept', type='terms', field='top_concept') conceptid_agg = Aggregation(name='conceptid', type='terms', field='conceptid') max_score_agg = MaxAgg(name='max_score', script='_score') top_concept_agg.add_aggregation(conceptid_agg) base_agg.add_aggregation(max_score_agg) base_agg.add_aggregation(top_concept_agg) base_agg.add_aggregation(nodegroupid_agg) query.add_aggregation(base_agg) ret[index] = [] results = query.search(index=index) for result in results['aggregations']['value_agg']['buckets']: if len(result['top_concept']['buckets']) > 0: for top_concept in result['top_concept']['buckets']: top_concept_id = top_concept['key'] top_concept_label = get_preflabel_from_conceptid(top_concept['key'], lang)['value'] for concept in top_concept['conceptid']['buckets']: ret[index].append({ 'type': 'concept', 'context': top_concept_id, 'context_label': top_concept_label, 'id': i, 'text': result['key'], 'value': concept['key'] }) i = i + 1 else: ret[index].append({ 'type': 'term', 'context': '', 'context_label': get_resource_model_label(result), 'id': i, 'text': result['key'], 'value': result['key'] }) i = i + 1 return JSONResponse(ret)
def delete(self, *args, **kwargs): se = SearchEngineFactory().create() request = kwargs.pop('request', None) provisional_edit_log_details = kwargs.pop('provisional_edit_log_details', None) for tile in self.tiles: tile.delete(*args, request=request, **kwargs) try: user = request.user user_is_reviewer = request.user.groups.filter(name='Resource Reviewer').exists() except AttributeError: #no user user = None if user_is_reviewer is True or self.user_owns_provisional(user): query = Query(se) bool_query = Bool() bool_query.filter(Terms(field='tileid', terms=[self.tileid])) query.add_query(bool_query) results = query.search(index='terms')['hits']['hits'] for result in results: se.delete(index='terms', id=result['_id']) self.__preDelete(request) self.save_edit( user=request.user, edit_type='tile delete', old_value=self.data, provisional_edit_log_details=provisional_edit_log_details) super(Tile, self).delete(*args, **kwargs) resource = Resource.objects.get(resourceinstanceid=self.resourceinstance.resourceinstanceid) resource.index() else: self.apply_provisional_edit(user, data={}, action='delete') super(Tile, self).save(*args, **kwargs)
def get_restricted_instances(user, search_engine=None, allresources=False): if allresources is False and user.is_superuser is True: return [] if allresources is True: restricted_group_instances = { perm["object_pk"] for perm in GroupObjectPermission.objects.filter(permission__codename="no_access_to_resourceinstance").values("object_pk") } restricted_user_instances = { perm["object_pk"] for perm in UserObjectPermission.objects.filter(permission__codename="no_access_to_resourceinstance").values("object_pk") } all_restricted_instances = list(restricted_group_instances | restricted_user_instances) return all_restricted_instances else: terms = Terms(field="permissions.users_with_no_access", terms=[str(user.id)]) query = Query(search_engine, start=0, limit=settings.SEARCH_RESULT_LIMIT) has_access = Bool() nested_term_filter = Nested(path="permissions", query=terms) has_access.must(nested_term_filter) query.add_query(has_access) results = query.search(index=RESOURCES_INDEX, scroll="1m") scroll_id = results["_scroll_id"] total = results["hits"]["total"]["value"] if total > settings.SEARCH_RESULT_LIMIT: pages = total // settings.SEARCH_RESULT_LIMIT for page in range(pages): results_scrolled = query.se.es.scroll(scroll_id=scroll_id, scroll="1m") results["hits"]["hits"] += results_scrolled["hits"]["hits"] restricted_ids = [res["_id"] for res in results["hits"]["hits"]] return restricted_ids
def delete(self, user={}, note=''): """ Deletes a single resource and any related indexed data """ se = SearchEngineFactory().create() related_resources = self.get_related_resources(lang="en-US", start=0, limit=1000) for rr in related_resources['resource_relationships']: models.ResourceXResource.objects.get(pk=rr['resourcexid']).delete() query = Query(se) bool_query = Bool() bool_query.filter( Terms(field='resourceinstanceid', terms=[self.resourceinstanceid])) query.add_query(bool_query) results = query.search(index='strings', doc_type='term')['hits']['hits'] for result in results: se.delete(index='strings', doc_type='term', id=result['_id']) se.delete(index='resource', doc_type=str(self.graph_id), id=self.resourceinstanceid) self.save_edit(edit_type='delete') super(Resource, self).delete()
def append_instance_permission_filter_dsl(request, search_results_object): if request.user.is_superuser is False: has_access = Bool() terms = Terms(field="permissions.users_with_no_access", terms=[str(request.user.id)]) has_access.must_not(terms) search_results_object["query"].add_query(has_access)
def delete(self, *args, **kwargs): se = SearchEngineFactory().create() request = kwargs.pop("request", None) provisional_edit_log_details = kwargs.pop("provisional_edit_log_details", None) for tile in self.tiles: tile.delete(*args, request=request, **kwargs) try: user = request.user user_is_reviewer = user_is_resource_reviewer(user) except AttributeError: # no user user = None user_is_reviewer = True if user_is_reviewer is True or self.user_owns_provisional(user): query = Query(se) bool_query = Bool() bool_query.filter(Terms(field="tileid", terms=[self.tileid])) query.add_query(bool_query) results = query.search(index="terms")["hits"]["hits"] for result in results: se.delete(index="terms", id=result["_id"]) self.__preDelete(request) self.save_edit( user=request.user, edit_type="tile delete", old_value=self.data, provisional_edit_log_details=provisional_edit_log_details ) super(Tile, self).delete(*args, **kwargs) resource = Resource.objects.get(resourceinstanceid=self.resourceinstance.resourceinstanceid) resource.index() else: self.apply_provisional_edit(user, data={}, action="delete") super(Tile, self).save(*args, **kwargs)
def get_preflabel_from_conceptid(conceptid, lang): ret = None default = { "category": "", "conceptid": "", "language": "", "value": "", "type": "", "id": "" } se = SearchEngineFactory().create() query = Query(se) terms = Terms(field='conceptid', terms=[conceptid]) match = Match(field='type', query='preflabel', type='phrase') query.add_filter(terms) query.add_query(match) preflabels = query.search(index='concept_labels')['hits']['hits'] for preflabel in preflabels: default = preflabel['_source'] # get the label in the preferred language, otherwise get the label in the default language if preflabel['_source']['language'] == lang: return preflabel['_source'] if preflabel['_source']['language'].split('-')[0] == lang.split( '-')[0]: ret = preflabel['_source'] if preflabel['_source'][ 'language'] == settings.LANGUAGE_CODE and ret == None: ret = preflabel['_source'] return default if ret == None else ret
def build_search_results_dsl(request): temporal_filters = JSONDeserializer().deserialize(request.GET.get('temporalFilter', None)) query = build_base_search_results_dsl(request) boolfilter = Bool() if 'filters' in temporal_filters: for temporal_filter in temporal_filters['filters']: terms = Terms(field='date_groups.conceptid', terms=temporal_filter['date_types__value']) boolfilter.must(terms) date_value = datetime.strptime(temporal_filter['date'], '%d/%m/%Y').isoformat() if temporal_filter['date_operators__value'] == '1': # equals query range = Range(field='date_groups.value', gte=date_value, lte=date_value) elif temporal_filter['date_operators__value'] == '0': # greater than query range = Range(field='date_groups.value', lt=date_value) elif temporal_filter['date_operators__value'] == '2': # less than query range = Range(field='date_groups.value', gt=date_value) if 'inverted' not in temporal_filters: temporal_filters['inverted'] = False if temporal_filters['inverted']: boolfilter.must_not(range) else: boolfilter.must(range) query.add_filter(boolfilter) return query
def create_nested_attribute_filter(self, doc_id, nodegroup_id, value_list): new_string_filter = Bool() new_string_filter.filter(Terms(field='strings.nodegroup_id', terms=[nodegroup_id])) for value in value_list: new_string_filter.should(Match(field='strings.string', query=value, type='phrase')) nested = Nested(path='strings', query=new_string_filter) return nested
def delete(self, user={}, note=""): """ Deletes a single resource and any related indexed data """ permit_deletion = False graph = models.GraphModel.objects.get(graphid=self.graph_id) if graph.isactive is False: message = _("This model is not yet active; unable to delete.") raise ModelInactiveError(message) if user != {}: user_is_reviewer = user_is_resource_reviewer(user) if user_is_reviewer is False: tiles = list(models.TileModel.objects.filter(resourceinstance=self)) resource_is_provisional = True if sum([len(t.data) for t in tiles]) == 0 else False if resource_is_provisional is True: permit_deletion = True else: permit_deletion = True else: permit_deletion = True if permit_deletion is True: related_resources = self.get_related_resources(lang="en-US", start=0, limit=1000, page=0) for rr in related_resources["resource_relationships"]: # delete any related resource entries, also reindex the resource that references this resource that's being deleted try: resourceXresource = models.ResourceXResource.objects.get(pk=rr["resourcexid"]) resource_to_reindex = ( resourceXresource.resourceinstanceidfrom_id if resourceXresource.resourceinstanceidto_id == self.resourceinstanceid else resourceXresource.resourceinstanceidto_id ) resourceXresource.delete(deletedResourceId=self.resourceinstanceid) res = Resource.objects.get(pk=resource_to_reindex) res.load_tiles() res.index() except ObjectDoesNotExist: se.delete(index=RESOURCE_RELATIONS_INDEX, id=rr["resourcexid"]) query = Query(se) bool_query = Bool() bool_query.filter(Terms(field="resourceinstanceid", terms=[self.resourceinstanceid])) query.add_query(bool_query) results = query.search(index=TERMS_INDEX)["hits"]["hits"] for result in results: se.delete(index=TERMS_INDEX, id=result["_id"]) se.delete(index=RESOURCES_INDEX, id=self.resourceinstanceid) try: self.save_edit(edit_type="delete", user=user, note=self.displayname) except: pass super(Resource, self).delete() return permit_deletion
def append_dsl(self, search_results_object, permitted_nodegroups, include_provisional): search_query = Bool() if include_provisional is not True: provisional_resource_filter = Bool() if include_provisional is False: provisional_resource_filter.filter( Terms(field='provisional_resource', terms=['false', 'partial'])) elif include_provisional is 'only provisional': provisional_resource_filter.filter( Terms(field='provisional_resource', terms=['true', 'partial'])) search_query.must(provisional_resource_filter) search_results_object['query'].add_query(search_query)
def append_dsl(self, search_results_object, permitted_nodegroups, include_provisional): search_query = Bool() if include_provisional is not True: provisional_resource_filter = Bool() if include_provisional is False: provisional_resource_filter.filter( Terms(field="provisional_resource", terms=["false", "partial"])) elif include_provisional == "only provisional": provisional_resource_filter.filter( Terms(field="provisional_resource", terms=["true", "partial"])) search_query.must(provisional_resource_filter) search_results_object["query"].add_query(search_query)
def append_null_search_filters(self, value, node, query, request): """ Appends the search query dsl to search for fields that have not been populated """ base_query = Bool() base_query.filter(Terms(field="graph_id", terms=[str(node.graph_id)])) null_query = Bool() data_exists_query = Exists(field="tiles.data.%s" % (str(node.pk))) nested_query = Nested(path="tiles", query=data_exists_query) null_query.must(nested_query) if value["op"] == "null": # search for tiles that don't exist exists_query = Bool() exists_query.must_not(null_query) base_query.should(exists_query) # search for tiles that do exist, but that have null or [] as values func_query = Dsl() func_query.dsl = { "function_score": { "min_score": 1, "query": {"match_all": {}}, "functions": [ { "script_score": { "script": { "source": """ int null_docs = 0; for(tile in params._source.tiles){ if(tile.data.containsKey(params.node_id)){ def val = tile.data.get(params.node_id); if (val == null || (val instanceof List && val.length==0)) { null_docs++; break; } } } return null_docs; """, "lang": "painless", "params": {"node_id": "%s" % (str(node.pk))}, } } } ], "score_mode": "max", "boost": 1, "boost_mode": "replace", } } base_query.should(func_query) elif value["op"] == "not_null": base_query.must(null_query) query.must(base_query)
def get_related_resources(self, lang='en-US', limit=1000, start=0): """ Returns an object that lists the related resources, the relationship types, and a reference to the current resource """ ret = { 'resource_instance': self, 'resource_relationships': [], 'related_resources': [] } se = SearchEngineFactory().create() query = Query(se, limit=limit, start=start) bool_filter = Bool() bool_filter.should( Terms(field='resourceinstanceidfrom', terms=self.resourceinstanceid)) bool_filter.should( Terms(field='resourceinstanceidto', terms=self.resourceinstanceid)) query.add_query(bool_filter) resource_relations = query.search(index='resource_relations', doc_type='all') ret['total'] = resource_relations['hits']['total'] instanceids = set() for relation in resource_relations['hits']['hits']: relation['_source']['preflabel'] = get_preflabel_from_valueid( relation['_source']['relationshiptype'], lang) ret['resource_relationships'].append(relation['_source']) instanceids.add(relation['_source']['resourceinstanceidto']) instanceids.add(relation['_source']['resourceinstanceidfrom']) if len(instanceids) > 0: instanceids.remove(str(self.resourceinstanceid)) related_resources = se.search(index='resource', doc_type='_all', id=list(instanceids)) if related_resources: for resource in related_resources['docs']: ret['related_resources'].append(resource['_source']) return ret
def delete_index(self, resourceinstanceid=None): """ Deletes all references to a resource from all indexes Keyword Arguments: resourceinstanceid -- the resource instance id to delete from related indexes, if supplied will use this over self.resourceinstanceid """ if resourceinstanceid is None: resourceinstanceid = self.resourceinstanceid resourceinstanceid = str(resourceinstanceid) # delete any related terms query = Query(se) bool_query = Bool() bool_query.filter(Terms(field="resourceinstanceid", terms=[resourceinstanceid])) query.add_query(bool_query) query.delete(index=TERMS_INDEX) # delete any related resource index entries query = Query(se) bool_query = Bool() bool_query.should(Terms(field="resourceinstanceidto", terms=[resourceinstanceid])) bool_query.should(Terms(field="resourceinstanceidfrom", terms=[resourceinstanceid])) query.add_query(bool_query) query.delete(index=RESOURCE_RELATIONS_INDEX) # reindex any related resources query = Query(se) bool_query = Bool() bool_query.filter(Nested(path="ids", query=Terms(field="ids.id", terms=[resourceinstanceid]))) query.add_query(bool_query) results = query.search(index=RESOURCES_INDEX)["hits"]["hits"] for result in results: res = Resource.objects.get(pk=result["_id"]) res.load_tiles() res.index() # delete resource index se.delete(index=RESOURCES_INDEX, id=resourceinstanceid)
def delete(self, user={}, note=''): """ Deletes a single resource and any related indexed data """ permit_deletion = False graph = models.GraphModel.objects.get(graphid=self.graph_id) if graph.isactive is False: message = _('This model is not yet active; unable to delete.') raise ModelInactiveError(message) if user != {}: user_is_reviewer = user.groups.filter( name='Resource Reviewer').exists() if user_is_reviewer is False: tiles = list( models.TileModel.objects.filter(resourceinstance=self)) resource_is_provisional = True if sum( [len(t.data) for t in tiles]) == 0 else False if resource_is_provisional is True: permit_deletion = True else: permit_deletion = True else: permit_deletion = True if permit_deletion is True: se = SearchEngineFactory().create() related_resources = self.get_related_resources(lang="en-US", start=0, limit=1000, page=0) for rr in related_resources['resource_relationships']: models.ResourceXResource.objects.get( pk=rr['resourcexid']).delete() query = Query(se) bool_query = Bool() bool_query.filter( Terms(field='resourceinstanceid', terms=[self.resourceinstanceid])) query.add_query(bool_query) results = query.search(index='terms')['hits']['hits'] for result in results: se.delete(index='terms', id=result['_id']) se.delete(index='resources', id=self.resourceinstanceid) self.save_edit(edit_type='delete', user=user, note=self.displayname) super(Resource, self).delete() return permit_deletion
def get_related_resources(resourceid, lang='en-US', limit=1000, start=0): ret = {'resource_relationships': [], 'related_resources': []} se = SearchEngineFactory().create() query = Query(se, limit=limit, start=start) query.add_filter(Terms(field='entityid1', terms=resourceid).dsl, operator='or') resource_relations = query.search(index='resource_relations', doc_type='all') ret['total'] = resource_relations['hits']['total'] for relation in resource_relations['hits']['hits']: ret['resource_relationships'].append(relation['_source']) return ret